Your Industry

Power Answer adapts to your business

Let's Talk

Uncategorized

Power Answer
Why Are My Google Ads Conversions Not Matching Actual Sales?

Why Are My Google Ads Conversions Not Matching Actual Sales?

Google Ads conversions do not match actual sales because a conversion is a recorded action, not a confirmed outcome. The platform counts a form submission, a phone call that rings past a length threshold, or a click on a tracked button, and it has no way of knowing whether that action turned into a booked job, a qualified opportunity, or a wrong number. Until the sale itself is fed back into the account, the conversion column measures activity volume and the revenue line measures reality, and the two will drift apart every single month.

That drift is not a bug in Google Ads. It is a data gap between the click and the closed deal, and it is usually caused by five specific, diagnosable things. Below is how each one inflates or distorts your numbers, and what to check in your own account.

Cause 1: Every Form Fill Counts as a Lead, Regardless of Quality

Google Ads treats a form submission as a completed conversion the moment the thank-you page fires or the submit event triggers. It cannot distinguish between a homeowner ready to book, a competitor pulling pricing, a job applicant, an existing customer looking for support, or a bot filling fields with garbage. In service categories with high-intent search terms, a meaningful portion of raw form volume is never a sellable opportunity at all.

The result is a conversion count that looks healthy while the sales pipeline stays thin. Worse, Smart Bidding is optimizing toward whatever produces those submissions, which means budget flows to the keywords, placements, and audiences that generate the most form activity, not the most revenue. Broad terms and information-seeking queries tend to win that contest.

Cause 2: Calls Are Counted at Ring, Not at a Qualified Outcome

By default, Google Ads counts a call conversion when the call exceeds a minimum duration, commonly set at 60 seconds. Sixty seconds is enough time for a caller to reach a receptionist, ask a question that was never a buying question, and hang up. It is also enough time for a robocall, a supplier, or a customer calling about an existing invoice to register as a fresh conversion.

Raising the threshold helps at the margin, but duration is a proxy for intent, not a measure of it. A 45-second call from a qualified buyer who says “can you come Thursday” is a real sale that gets discarded. A three-minute call from someone outside your service area is counted as a win. Length alone cannot tell you which is which, which is why call outcome data matters more than call volume.

Cause 3: One Person Is Being Counted as Several Conversions

Real buyers rarely convert once. They submit a form, call the next morning, click a different ad that evening, and call again to confirm scheduling. With conversion counting set to “every” instead of “one,” and with click-through conversion windows that can extend up to 90 days, a single customer can register three or four conversions across two or three campaigns.

Check these four settings before you assume the platform is broken:

  • Counting setting: “Every” is appropriate for ecommerce transactions, “one” for lead generation actions.
  • Conversion window: A long window inflates the count of a slow sales cycle without inflating revenue.
  • Duplicate actions: A thank-you page conversion plus a form-submit event plus a click-to-call event can triple-count the same behavior.
  • Primary vs. secondary actions: Only actions marked primary should be feeding bidding. Micro-conversions left as primary skew everything downstream.

Cause 4: Missed and Abandoned Calls Never Enter the Record

This is the failure that runs in the opposite direction, and it hurts more than over-counting. A call that rings out, hits voicemail, dies on hold, or lands after hours generally never reaches the duration threshold, so it never becomes a conversion. The click was paid for, the intent was real, and the account shows nothing.

Over a quarter, that pattern quietly teaches the bidding algorithm that your highest-intent hours and highest-intent keywords are underperformers, because the calls they generated were the ones nobody answered. Trade and home services advertisers feel this most acutely, since demand arrives in bursts around weather events, early mornings, and evenings when phone coverage is thinnest.

Cause 5: The Sale Happens Offline and Never Goes Back to Google

Most service businesses close revenue days or weeks after the click, in a CRM, on a second call, or on site. Unless that outcome is imported back into Google Ads, the platform never learns which clicks produced money. Offline conversion import exists precisely for this: the GCLID captured at the click is stored with the lead, then uploaded later with a status and a value, generally within 90 days of the original click. Enhanced conversions for leads does the equivalent using hashed contact details instead of a stored click ID.

Without that upload, value-based bidding strategies such as target ROAS have nothing real to optimize against, so they optimize against your best guess of a static conversion value instead.

How to Reconcile Conversions Against Sales, Step by Step

  1. Pick one closed month and export every Google Ads conversion with date, campaign, and conversion action.
  2. Export every lead your team actually worked in that month, including calls that were never answered.
  3. Deduplicate by phone number and email so repeat touches collapse into one contact.
  4. Tag each contact with a real outcome: qualified, unqualified, out of area, existing customer, spam, sold.
  5. Compare four numbers side by side: platform conversions, unique contacts, qualified contacts, and closed revenue. The gaps between them tell you which of the five causes above is dominant in your account.
  6. Fix the counting settings first, then start feeding qualified outcomes and revenue back into the platform.

Closing the Loop With Call Recording, Transcription, and Contact History

The durable fix is to make every interaction produce a verdict, not just a timestamp. Call recording paired with transcription gives you the actual words of the conversation. AI summaries turn each of those transcripts into a short, structured read on what the caller wanted, whether they were in the service area, and whether the call ended in a booking, so qualification stops depending on someone remembering to update a spreadsheet. Recent platform updates now load call details faster and attach AI-written summaries directly to the call record, and contact profiles carry status tracking, automatic tagging based on call activity, and manual notes alongside the full call and form-submission history for that person.

Contact-level history is what makes deduplication and attribution honest. When four touches roll up to one person with one status and one revenue figure, you can push that single qualified outcome back into Google Ads as a conversion with a real value. At that point the bid algorithm is training on revenue rather than raw actions, and the keywords that generate answered, qualified, closed calls start absorbing the budget that broad, form-heavy terms were consuming.

Frequently Asked Questions

Should Google Ads Conversions Ever Match Sales Exactly?

No. Even with clean tracking, modeled conversions, cross-device journeys, and consent-based measurement mean the platform reports an estimate. The goal is a stable, explainable relationship between conversions and revenue, not a perfect one-to-one match.

Why Are Conversions Up but Revenue Flat?

Almost always because volume grew in a segment that does not convert to sales: broader match types, low-intent queries, or spam. Look at qualified contacts per campaign rather than conversions per campaign, and the flat revenue usually explains itself.

Does Raising the Call Length Threshold Fix Bad Call Conversions?

It reduces obvious noise, but it also discards fast, high-intent bookings and still counts long, unqualified calls. Outcome-based qualification from transcripts and summaries is the more accurate control.

What Should Agencies Report Instead of Raw Conversions?

Report unique contacts, qualified rate, answered-call rate, and closed revenue by campaign. Those four figures survive client scrutiny in a way a conversion count never does, and they are the metrics an agency reporting workflow should be built around.

A conversion count that outpaces revenue is a measurement problem with a specific cause, and the diagnostic above will usually isolate it in an afternoon. Power Answer publishes ongoing analysis of call data, lead quality, and attribution accuracy for teams working through exactly this gap between reported conversions and money in the bank.